Industrial production of cellulose is an energy intensive process. Businesses aim to utilized as much input energy, materials as possible while minimizing the waste as well. In addition to the main product – cellulose and large amounts of organically polluted water which is subsequently treated in waste water treatment plants. This PhD thesis deals with formation of suitable mixing formula for sludge from cellulose production and available materials (waste) from close neighborhood of Biocel Paskov a.s. so that well balanced cofermentation products are achieved. This mainly involves grass from lawn maintenance, grass ensilage, potato peels, and leftovers from vegetable processing. Fermentation processes (both mesophilic and thermophilic) of prepared fermentation products were conducted in semi-continuous and continuous laboratory fermentation units. Reduction of organic mass depending on residence time was closely observed as well as production and quality of biogas along with quality of output digestate. Course of process behaviour under controlled pH was tested. Results of particular tests were integrated into graphs. Conclusion of the thesis presents balance scheme drawn for selected variants, and design of real fermentation station with individual buildings, operations, basic machinery and equipment description including investment costs. Financial calculation and expected investment return was conducted in two variants – without subsidies and including subsidies from Operational Programme Renewable Sources of Energy.

Östersjön är idag det system med störst utbredning av syrefattiga eller syrefria bottnar i hela världen. Problemet har uppstått som följd av den massiva näringsbelastning från de kringliggande länderna som pågått sedan industrialiseringen intensifierades på 1900-talet. I och med omställningen till förnybar energi planeras framtida vindkraftsparker att byggas i Östersjön. Överflödig energi planeras att användas till havsbaserad vätgasproduktion som sedan forslas in till land med rör. Vid storskalig vätgasproduktion produceras som biprodukt enorma mängder syrgas vilket ger helt nya förutsättningar för artificiell syresättning av Östersjöns syrefria bottnar. Tre övergripande tekniker för att artificiellt syresätta bottenvatten med syrgas har identifierats på 21 platser runt om i världen: Downflow Bubble Contact System (DBCS), Side Stream Pumping System (SSPS) och Direct Oxygen Injection System (DOIS). Alla tre tekniker har med goda resultat förhöjt syrgaskoncentrationen i bottenvatten från hypoxiskt (< 2 mg/l) till oxiskt (> 2 mg/l). I flera fall har syresättningen bidragit till reduktion av näringsämnen och/eller utökat habitat för fiskar och bottendjur. En pilotanläggning i fjärden Slätbaken skulle ge bättre förståelse för hur ett system skulle kunna utformas i andra delar av Östersjön. En massbalansmodell konstruerades över Slätbaken och resultatet var att pilotanläggningen bör ha en kapacitet på 8 ton O2/dygn för att med säkerhet kunna höja syrgaskoncentrationen till > 4 mg/l. Detta skulle ge en lämplig säkerhetsmarginal för den lokala bottennära eller bottenlevande faunan med avseende på syrgashalt. / As of today, the Baltic Sea is the system with the globally largest expansion of hypoxic or anoxic bottom waters, so-called “dead zones”. It is a consequence of the massive load of nutrients from the surrounding countries that started with the intensified industrialization in the start of the 20th century. With the transformation towards renewable energy, wind parks are planned to be built in the Baltic Sea on a substantial scale. Thereby, superfluous energy is planned to be used for sea-based production of hydrogen gas that can be piped to land. Large scale hydrogen gas production would generate enormous amounts of oxygen gas as a byproduct, which gives a theoretical opportunity to artificially oxygenate the dead zones of the Baltic Sea. Three main techniques to artificially oxygenate bottom waters were identified in 21 freshwater systems around the world: Downflow Bubble Contact System (DBCS), Side Stream Pumping System (SSPS) and Direct Oxygen Injection System (DOIS). All three techniques have successfully raised the concentration of dissolved oxygen from hypoxic (< 2 mg/l) to oxic (> 2 mg/l). In many cases, the oxygenation contributed to reduction of dissolved nutrients in the waters and/or enlarged habitats for fish and benthic fauna. A pilot plant for oxygenation in the bay Slätbaken would contribute to better understanding of how a future oxygenation system could be constructed in other parts of the Baltic Sea. A mass balance model was constructed, representing oxygen content and transport in Slätbaken. Results indicated that establishing a pilot plant would need a capacity of 8 tons O2/day to guarantee a dissolved oxygen concentration of > 4 mg/l, which would provide local fauna with a suitable safety margin regarding dissolved oxygen concentrations.

The satellite gravimetry mission Gravity Record And Climate Experiment (GRACE), which was operational from 2002 to 2017, and its follow-on mission GRACE-Follow-On (GRACE-FO), which has been active since 2018, revolutionized the observation of temporal changes of the Earth's gravitational field. The measurement data from these missions enable the nuanced quantification of mass redistributions on Earth. Water redistributions between continents and oceans caused by climate change are of particular research interest because of their relevance for mankind. These are, for example, the ice mass changes (IMC) of the ice sheets in Antarctica and Greenland, which this work focuses on. IMC estimates derived from satellite gravimetry data, like from other quantification methods, confirm that both the Greenland Ice Sheet (GIS) and the Antarctic Ice Sheet (AIS) have been losing mass over the last two decades. However, these estimates are subject to large uncertainties, which is particularly the case for the AIS. If the mass balance is obtained from gravimetric observations, a major source of uncertainty is the consideration of effects due to glacial isostatic adjustment (GIA). The uncertainty of the present-day gravitational field changes caused by the isostatic adjustment of the solid Earth to IMC during the last centuries and millennia propagates into estimates of the recent IMC. According to results of the Ice sheet Mass Balance Inter-comparison Exercise (IMBIE), the spread of different modelling results predicting the GIA-induced mass effect in Antarctica is almost as large as the estimated rate of the IMC itself. In Greenland, the spread of the mass effect from different GIA modelling results is approximately 20 % of the rate of IMC. Alternatively, the IMC can be determined using surface elevation changes derived from satellite altimetry observations. In this case, any GIA error hardly affects the results, but there is a significant source of uncertainty in the conversion of volume changes into mass changes. It is possible to combine data from satellite gravimetry and satellite altimetry to jointly estimate IMC and GIA mass effects, e.g. by solving an inverse problem (joint data inversion). This is an alternative to the use of GIA modelling results in processing satellite gravimetry data. Results from data combination methods are not only a means to an end to improve the estimation of IMC. They also can contribute to answer geodynamic questions. However, previous estimation strategies for combining satellite gravimetry and satellite altimetry data are subject to some limitations. Many approaches only allow to estimate GIA in a regional framework and not in global framework. Other approaches strongly depend on a priori information from geophysical modelling which are subject to large uncertainties. Furthermore, limitations are due to processing choices, e.g. the use of deterministic parameters over defined time intervals or, e.g. due to the consideration of errors in the applied data sets. This work investigates advancements of data combination methods that allow to quantify IMC and present-day GIA effects. Specifically, the approaches investigated here combine measured gravitational field changes from satellite gravimetry, measured surface elevation changes from satellite altimetry, modelled surface mass balances from regional climate modelling, and modelled firn thickness changes from firn modelling. This cumulative dissertation comprises three publications that investigated three aspects of data combination approaches. The first publication analysed a regional combination approach in Antarctica and results therein demonstrated a significant dependence of the estimated GIA effect on the input data sets and applied processing choices. A bias correction can significantly reduce an initial bias in the determined GIA effect associated to the spherical harmonic coefficients of degree-1 and c₂₀. However, this bias correction regionally constrains the GIA estimate and prevents to implement such an approach in a global framework. The second publication infers long-term mass trends with their temporal changes jointly observed from satellite gravimetry and satellite altimetry data. To do so, a state-space filtering framework was applied to the data sets allowing to estimate temporal changes of the parameters over time while accounting for temporal correlation of short-term fluctuations. Thereby, an accelerating ice-dynamically induced ice mass loss is found for drainage basins in West Antarctica. In contrast, the temporal variability of long-term trends in East Antarctica is low. Noteworthy, the trends in Dronning Maud Land and Enderby Land are positive. The third publication presents a global approach to jointly estimate IMC, GIA effects and firn thickness changes, while accounting for spatial error covariances of the input data sets. The intention of the utilized GIA parametrization in Antarctica is to spatially resolve GIA effects that were not predicted by GIA models. Simulation experiments demonstrated the feasibility of the approach under the presence of realistic limitations of satellite observations and model products. This framework paper also reports a first application of the inversion method of Publication~3 to real data. The focus of this application is on Antarctcia over the time interval January 2011 to December 2020. Results for the AIS are: (i) an IMC of (−150 ± 5) Gt a⁻¹, (ii) a change of the firn air content of (40 ± 5) km³ a⁻¹, and (iii) an integrated GIA-induced mass effect of (72 ± 4) Gt a⁻¹. These results are promising with regard to the application of this methodology, as they are similar to previously published estimates. But they are estimated in a globally consistent framework and without applying conventional filtering strategies. Future work should further improve the methodology and eventually implement it in a global inversion framework that allows to jointly estimate all sea-level contributions.:1 Introduction 2 Processes over ice sheets inducing changes in Earth’s gravity and geometry 3 Data sets 4 Data combinations over ice sheets 5 Publications 6 Inversion of real data for glacial isostatic adjustment and ice mass changes in Antarctica 7 Outlook 8 Conclusions / Die Satellitengravimetriemission Gravity Record And Climate Experiment (GRACE), die von 2002 bis 2017 aktiv war, sowie die seit 2018 aktive Nachfolgemission GRACE-Follow-On (GRACE-FO) revolutionierten die Beobachtung zeitlicher Änderungen des Gravitationsfeldes der Erde. Die Messdaten dieser Missionen ermöglichen die differenzierte Quantifizierung von Massenumverteilungen auf der Erde. Von besonderen Forschungsinteresse, aufgrund ihrer Relevanz für die Menschheit, sind dabei durch den Klimawandel verursachte Umverteilungen von Wasser zwischen den Kontinenten und dem Ozean. Das sind beispielsweise die Eismassenänderungen der Eisschilde in Antarktika sowie Grönland, die im Fokus dieser Arbeit stehen. Aus Messdaten der Satellitengravimetrie ermittelte Eismassenänderungen bestätigen, wie auch andere Quantifizierungsmethoden, dass der Grönländische Eisschild sowie der Antarktische Eisschild während der letzten zwei Jahrzehnte an Masse verloren haben. Allerdings sind diese Schätzungen mit großen Unsicherheiten behaftet, was insbesondere auf den Antarktischen Eisschild zutrifft. Wird die Massenbilanz mit gravimetrischen Beobachtungen ermittelt, ist eine wesentliche Quelle für die Unsicherheit die Berücksichtigung der Effekte aufgrund des glazial-isostatischen Ausgleichs (GIA). Die Unsicherheit über die gegenwärtigen Änderungen des Gravitationsfeldes, aufgrund des isostatischen Ausgleichs der festen Erde an Eismassenänderungen während der letzten Jahrhunderte und Jahrtausende, pflanzt sich in die Schätzung rezenter Massenänderungen fort. Laut Ergebnissen von vergleichenden Untersuchungen zu Eisschildmassenbilanzen (Ice sheet Mass Balance Inter-comparison Exercise, IMBIE) ist in Antarktika die Bandbreite unterschiedlicher Modellierungen des GIA-induzierten Masseneffekts fast so groß wie die ermittelte Rate der Eismassenänderung selbst. In Grönland beträgt die Bandbreite des Masseneffekts unterschiedlicher GIA-Modellierungen ungefähr 20 % der Eismassenänderungsrate. Alternativ lassen sich die Eismassenänderungen mittels Oberflächenhöhenänderungen bestimmen, die aus Beobachtungen der Satellitenaltimetrie abgeleitet werden. Dabei beeinflussen GIA Fehler die Ergebnisse kaum, allerdings besteht dabei eine wesentliche Quelle der Unsicherheit bei der Konversion von Volumenänderungen in Massenänderungen. Es besteht die Möglichkeit, Daten der Satellitengravimetrie sowie der Satellitenaltimetrie zu kombinieren und somit die Eismassenänderungen sowie GIA-Masseneffekte gemeinsam zu bestimmen, z. B. als Lösung eines inversen Problems (gemeinsame Dateninversion). Dies ist eine Alternative zur Verwendung von Ergebnissen der GIA-Modellierung in der Datenprozessierung der Satellitengravimetrie. Ergebnisse von Datenkombinationsmethoden sind dabei nicht nur ein Mittel zum Zweck, um die Schätzung von Eismassenänderungen zu verbessern. Sie können auch zur Beantwortung geodynamischer Fragestellungen beitragen. Allerdings unterliegen bisherige Schätzverfahren, die Daten der Satellitengravimetrie und Satellitenaltimetrie kombinieren, Limitierungen. Viele Ansätze ermöglichen die GIA Schätzungen nur in einem regionalen Rahmen und nicht in einem globalen Rahmen. Andere Ansätze hängen stark von Vorinformationen der geophysikalischen Modellierung ab, die aber große Unsicherheiten aufweisen. Außerdem ergeben sich Limitierungen durch gewählte Prozessierungsentscheidungen, wie z. B. durch die Verwendung deterministischer Parameter über definierte Zeitintervalle oder z. B. durch die Berücksichtigungen der Fehler der verwendeten Datensätze. Diese Arbeit untersucht Weiterentwicklungen von Datenkombinationsmethoden, welche die Quantifizierung von Eismassenänderungen und des gegenwärtigen GIA-induzierten Masseneffekts ermöglichen. Konkret kombinieren die hier untersuchten Ansätze: gemessene Gravitationsfeldänderungen der Satellitengravimetrie, gemessene Oberflächenhöhenänderungen der Satellitenaltimetrie, modellierte Oberflächenmassenbilanzen sowie modellierte Firndickenänderungen der regionalen Klimamodellierung. Diese kumulative Dissertation umfasst drei Publikationen, die drei Aspekte von Datenkombinationsansätzen untersuchten. Die erste Publikation analysierte einen regionalen Kombinationsansatzes in Antarktika und die Ergebnisse zeigten eine bedeutende Abhängigkeit des ermittelten GIA-Effekts von den verwendeten Eingangsdatensätzen und Prozessierungsentscheidungen. Ein ursprünglicher Bias im ermittelten GIA-Effekt, aufgrund der sphärisch-harmonischen Koeffizienten vom Grad-1 sowie c₂₀, kann durch eine Biaskorrektur erheblich reduziert werden. Dadurch sind die GIA-Schätzungen allerdings regional beschränkt und es wird verhindert, dass ein solcher Ansatz in einem globalen Rahmen implementiert werden kann. Die zweite Publikation ermittelt Langzeitmassentrends zusammen mit deren zeitlichen Änderungen, die von der Satellitengravimetrie und Satellitenaltimetrie gemeinsam beobachtet werden. Hierfür wurde ein Zustandsraumfilterverfahren auf die Datensätze angewandt, das es ermöglicht, die zeitlichen Veränderungen der Parameter über die Zeit zu bestimmen, unter der Berücksichtigung zeitlicher Korrelation kurzfristiger Fluktuationen. Dabei zeigt sich für Abflussbecken in der Westantarktis ein sich beschleunigender eisdynamisch induzierter Eismassenverlust. Dagegen ist die zeitliche Variabilität der Langzeittrends in der Ostantarktis gering. Bemerkenswert ist, dass die Trends im Dronning Maud Land und Enderby Land positiv sind. Die dritte Publikation präsentiert einen globalen Ansatz, der die gemeinsame Schätzung von Eismassenänderung, der GIA-Effekte sowie Änderungen der Firndicke ermöglicht, unter der Berücksichtigung räumlicher Fehlerkovarianzen. Bei der Wahl der GIA-Parametrisierung in Antarktika wurde die Intention verfolgt, GIA-Effekte räumlich aufzulösen, die bisher nicht von GIA-Modellen vorhergesagt wurden. Mit Simulationsexperimenten konnte die Machbarkeit des Ansatzes unter realistischer Limitierungen der Satelliten- und Modellprodukte demonstriert werden. Diese Rahmenschrift präsentiert auch eine erste Anwendung der Inversionsmethode aus Publikation 3 unter Verwendung echter Daten. Der Fokus dieser Anwendung liegt auf Antarktika über das Zeitintervall Januar 2011 bis Dezember 2020. Ergebnisse für den Antarktischen Eisschild sind: (i) eine Eismassenänderung von (−150 ± 5) Gt a⁻¹, (ii) eine Änderung des Luftgehalts der Firnschicht von (40 ± 5) km³ a⁻¹ und (iii) ein integrierter GIA-induzierter Masseneffekt von (72 ± 4) Gt a⁻¹. Diese Ergebnisse sind vielversprechend mit Hinblick auf die Anwendbarkeit der Methode, da sie vergleichbar zu bereits publizierten Ergebnissen sind. Dabei wurden sie in einem global-konsistenten Rahmen ohne die Anwendung konventioneller Filterungen ermittelt. Im Zuge zukünftigen Arbeiten soll die Methodik weiter verbessert werden und schließlich in einem globalen Inversionsrahmen implementiert werden, der die Bestimmung aller Meeresspiegelbeiträge gemeinsam ermöglicht.:1 Introduction 2 Processes over ice sheets inducing changes in Earth’s gravity and geometry 3 Data sets 4 Data combinations over ice sheets 5 Publications 6 Inversion of real data for glacial isostatic adjustment and ice mass changes in Antarctica 7 Outlook 8 Conclusions

Mit der Mission des Ice, Cloud and Land Elevation Satellite (ICESat) gelangte erstmals ein Laseraltimetersystem in einen erdgebundenen Orbit. Die vorliegende Arbeit verdeutlicht anhand von drei verschiedenen Anwendungen das Potenzial dieser Altimeterdaten zur Überwachung des Antarktischen und des Grönländischen Eisschilds. Beide Schilde bilden ein Schlüsselglied im globalen Klimasystem der Erde. In einem ersten Hauptabschnitt werden die ICESat-Altimeterdaten für das Gebiet des Lake Vostok, des größten Vertreters subglazialer Seen in der Antarktis, untersucht. Dieses Gebiet eignet sich durch die Höhenstabilität des über dem See liegenden Eisschilds insbesondere als Validierungsgebiet für Altimeterdaten. Diese werden hinsichtlich der zwischen den Lasern auftretenden Offsets umfassend analysiert. Die ermittelten Offsets variieren in einem Bereich zwischen -7.5 und +13.9 cm und erreichen damit die angestrebte Messgenauigkeit der Mission. Im Hinblick auf eine Bestimmung von zeitlich linearen Höhenänderungen der Eisschilde stellen sie den größten genauigkeits-limitierenden Faktor dar. Aus den um die Offsets korrigierten Altimeterdaten wird ein rasterförmiges Topographiemodell der Eisoberfläche erstellt. Dieses wird umfassend untersucht. Im Anschluss werden glaziologische Anwendungen vorgestellt, für welche das Topographiemodell eine zentrale Grundlage bildet. Unter anderem erfolgt in der Kombination mit Eisdicken- und Geoidinformationen der Nachweis, dass sich das Eis über dem See im hydrostatischen Gleichgewicht befindet. Im Zuge dieser Untersuchung wird aber auch deutlich, dass an einigen Stellen des Sees das Gleichgewicht verletzt wird. Mögliche Ursachen hierfür werden näher untersucht und eingehend diskutiert. Für den Grönländischen Eisschild erfolgt die Analyse der um die Laseroffsets korrigierten Altimeterdaten zur Ableitung zeitlich linearer Höhenänderungen. Die methodische Basis hierfür bildet eine Wiederholspuranalyse der Altimeterdaten. Zur Minimierung des Einflusses der lokalen Topographie und zur besseren Separation der saisonalen Höhenvariation werden die korrespondierenden Altimetermessungen entlang der Referenzspuren an ein drei-komponentiges mathematisches Modell durch Ausgleichung bestmöglich angepasst. Die für den ICESat-Missionszeitraum bestimmte mittlere Höhenrate des Eisschilds beträgt -13.0±0.5 cm/a. Die stärkste Höhenabnahme verzeichnet der Eisschild in den westlichen und südöstlichen küstennahen Randbereichen. Unter Verwendung der Eisdichte für die Volumen-Massen-Umrechnung entspricht dies einer Massenänderung von -209.5±35.6 Gt/a. Dies entspricht einem eustatischen Meeresspiegelanstieg von +0.6±0.1 mm/a. In einer dritten Anwendung werden die ICESat-Altimeterdaten über dem Amery-Schelfeises untersucht. Es wird eine Methodik vorgestellt, welche auf der Kreuzkorrelation von Höhenprofilen verschiedener Epochen beruht und zur Ableitung von Fließgeschwindigkeiten des Schelfeises dient. Der entwickelte Ansatz wird auf die ICESat-Referenzspur 49 angewendet. Sie verläuft entlang der zentralen Achse des Schelfeises. Im Bereich zwischen -71.6° und -70.1° Breite wächst die Fließgeschwindigkeit von +0.83±0.09 m/d auf +1.02±0.06 m/d an. Das Ergebnis steht im Einklang mit einem unabhängigen Geschwindigkeitsmodell, welches zur Validierung herangezogen wurde. / The Ice, Cloud and Land Elevation Satellite (ICESat) was the first Earth-orbiting laser altimeter mission in space. The following work is dedicated to the ICESat-altimetry data in order to demonstrate their full potential for the investigation of glaciological implications in polar regions. The primary science objective of the mission was to focus on the mass balances of the Greenland Ice Sheet and the Antarctic Ice Sheet. Both of them play a key role in the Earth's climate system. Firstly, the ICESat elevation profiles covering the Lake Vostok region are analysed in more detail. The Lake Vostok is the largest known subglacial lake in Antarctica to date. Due to a fast and strong degradation of the laser energy, the ICESat elevation measurements are affected by offsets. The estimated offsets between the laser operational periods vary between -7.5 und +13.9 cm. Therefore, they can't be neglected in the view of precise mass change determinations for ice sheets. In addition, a Digital Elevation Model (DEM) of the ice surface topography is generated on the basis of the adjusted elevation profiles. The DEM is analysed in more detail. Furthermore, the DEM forms the basis for the investigation of glaciological implications. In combination with an ice-thickness model and a regional geoid model the hydrostatic equilibrium condition is evaluated. It turns out, that the ice sheet covering the lake fulfils the hydrostatic equilibrium condition within ±1 m for large parts of the lake. Beside this, positive and negative deviations are found in the northern and southern part of the lake. Secondly, ice surface height changes and their temporal variations are inferred for the Greenland ice sheet. This investigation is based on a refined repeat-track analysis in order to exploit the full potential of ICESat's altimetry data. To reduce the influence of the local topography corresponding measurements along the track are fitted to a mathematical model, consisting of three components. For the entire ice sheet a mean surface height trend of -13.0±0.5 cm/yr is determined. The largest changes are identified at the coastal margins of the ice sheet. Using the ice surface height changes long-term volume- and mass-change rates are inferred. For this purpose the density of pure ice is used for the volume-mass-conversion. The overall long-term mass change rate amounts to -209.5±35.6 Gt/yr. This is equivalent to an eustatic sea level rise of +0.6±0.1 mm/yr. A third approach analyses ICESat elevation profiles over the Amery ice shelf. The method is based on a cross-correlation analysis of different ICESat repeat cycle in order to determine the ice flow velocity along the track. This method is applied to reference track 49. The investigation reveals that between 71.7° S and 70.1° S along the reference track, the ice-flow velocity increases from about +0.83±0.09 m/d to +1.02±0.06 m/d. These results are in general good agreement with velocities derived from an independent velocity field.

In the last decades, the increasing number of new generation satellite images characterized by a better spectral, spatial and temporal resolution with respect to the past has provided unprecedented source of information for monitoring climate changes.To exploit this wealth of data, powerful and automatic methods to analyze remote sensing images need to be implemented. Accordingly, the objective of this thesis is to develop advanced methods for the analysis of multitemporal multispectral remote sensing images to support climate change applications. The thesis is divided into two main parts and provides four novel contributions to the state-of-the-art. In the first part of the thesis, we exploit multitemporal and multispectral remote sensing data for accurately monitoring two essential climate variables. The first contribution presents a method to improve the estimation of the glacier mass balance provided by physically-based models. Unlike most of the literature approaches, this method integrates together physically-based models, remote sensing data and in-situ measurements to achieve an accurate and comprehensive glacier mass balance estimation. The second contribution addresses the land cover mapping for monitoring climate change at high spatial resolution. Within this work, we developed two processing chains: one for the production of a recent (2019) static high resolution (10 m) land cover map at subcontinental scale, and the other for the production of a long-term record of regional high resolution (30 m) land cover maps. The second part of this thesis addresses the common challenges faced while performing the analysis of multitemporal multispectral remote sensing data. In this context, the third contribution deals with the multispectral images cloud occlusions problem. Differently from the literature, instead of performing computationally expensive cloud restoration techniques, we study the robustness of deep learning architectures such as Long Short Term Memory classifier to cloud cover. Finally, we address the problem of the large scale training set definition for multispectral data classification. To this aim, we propose an approach that leverages on available low resolution land cover maps and domain adaptation techniques to provide representative training sets at large scale. The proposed methods have been tested on Sentinel-2 and Landsat 5, 7, 8 multispectral images. Qualitative and quantitative experimental results confirm the effectiveness of the methods proposed in this thesis.

[pt] Os modelos de balanço de massa servem como ferramenta para compreender e/ou quantificar o comportamento e destino de compostos químicos no meio ambiente, através dos diferentes compartimentos ambientais. Estes têm como finalidade principal a obtenção de um panorama geral e observação das conexões entre diversos fatores que impactam na concentração, no transporte e na transformação de compostos químicos. A Lagoa Rodrigo de Freitas e o Complexo Lagunar de Jacarepaguá, no Rio de Janeiro, são regiões costeiras que funcionam como bacias de acumulação sofrendo constantemente com problemas relacionados à poluição, por estarem situadas em regiões de alta urbanização, densidade demográfica e fluxo de veículos. Trabalhos de monitoramento realizados nas regiões apontam que as lagoas apresentam concentrações elevadas de metais, que quando liberados no meio ambiente tendem a se distribuir e acumular em diferentes compartimentos, bióticos e abióticos. Estes contaminantes apresentam persistência ambiental, estão frequentemente biodisponíveis no compartimento aquático e muitas vezes levam a efeitos tóxicos, tanto para a biota exposta quanto para os humanos. Todavia algumas regiões são de difícil monitoramento devido a questões logísticas e financeiras, e o uso desse tipo de modelagem vêm sendo empregada como uma ferramenta interessante. Neste contexto, o presente trabalho descreve os resultados da modelagem ambiental, tendo a aquivalência como critério de equilíbrio, na análise do destino e transporte de diferentes metais nas duas regiões supracitadas, através da estimativa e discussão das taxas de fluxos, regiões de acúmulo, variações locais e graus de complexidade dos modelos. Os resultados indicam que, em ambos os casos, houve um bom ajuste dos modelos para a realidade local. Na Lagoa Rodrigo de Freitas a maior remoção de metais ocorreu através do ar (89,88 por cento). Água e sedimento (49,11 por cento) foram as regiões que mais trocaram metais, sendo também a poeira urbana uma importante fonte de troca. Sedimento, poeira urbana e emissões diretas no meio aquático contribuíram mais significativamente para o transporte de metais para a água (maior 75 por cento). Na Lagoa de Jacarepaguá a maior remoção de metais ocorreu através do ar (49,36 por cento) e soterramento (28,42 por cento). A advecção do ar (49,39 por cento) e da água (44,68 por cento) foram as maiores vias de remoção de metais na Lagoa da Tijuca, enquanto na Lagoa de Marapendi a remoção se deu quase exclusivamente através da advecção no ar. A advecção da água e o sedimento foram as principais fontes de metais para as águas das Lagoas de Jacarepaguá e Tijuca, enquanto o ar foi o responsável por grande parte desse aporte na Lagoa de Marapendi, sendo justificado pela configuração espacial das lagoas e as premissas adotadas no modelo. As regiões com maior tendência de acúmulo de metais foram o sedimento, o solo e a poeira urbana. O maior grau de complexidade do modelo em comparação a um modelo mais simplificado e a maior quantidade/confiabilidade dos dados ambientais aumentaram a precisão das predições do modelo. Estes resultados certificam o uso desse tipo de modelagem como ferramenta auxiliar no controle, classificação e avaliação de risco e tomada de decisões pelos órgãos competentes. / [en] Mass balance models serve as a tool to understand and/or quantify the behavior and fate of chemical compounds in the environment, across different environmental compartments. These have as main purpose to obtain an overview and observation of the connections between different factors that impact the concentration, transport, and transformation of chemical compounds. The Rodrigo de Freitas lagoon and the Jacarepaguá lagoonal complex, in Rio de Janeiro, are coastal regions that function as accumulation basins, constantly suffering from problems related to pollution, as they are located in regions of high urbanization, demographic density and vehicles flow. Monitoring work carried out in the regions indicates that the lagoons have high concentrations of metals, which when released into the environment tend to be distributed and accumulated in different compartments, biotic and abiotic. These contaminants have environmental persistence, are often bioavailable in the aquatic compartment, and often lead to toxic effects, both for exposed biota and for humans. However, some regions are difficult to monitor due to logistical and financial issues, and the use of this type of modeling has been used as an interesting tool. In this context, the present work describes the results of the environmental modeling, having aquivalence as a balance criterion, in the analysis of the destination and transport of different metals in the two aforementioned regions, through the estimation and discussion of the flow rates, accumulation regions, local variations and degrees of complexity of the models. The results indicate that, in both cases, there was a good adjustment of the models to the local reality. In Rodrigo de Freitas lagoon, the highest removal of metals occurred through air (89.88 per cent). Water and sediment (49.11 per cent) were the regions that most exchanged metals, with road dust also being and important source of exchange. Sediment, road dust, and direct emissions into aquatic environment contributed most significantly to the transport of metals into water (more than 75 per cent). In the Jacarepaguá lagoon, the highest metal removal occurred through air (49.36 per cent) and burial (28.42 per cent). Air (49.39 per cent) and water (44.68 per cent) advection were the major means of metal removal in the Tijuca lagoon, while in the Marapendi lagoon removal occurred almost exclusively through air advection. The advection in water and sediment were the main sources of metals for the waters of the Jacarepaguá and Tijuca lagoons, while the air was responsible for a large part of this contribution in the Marapendi lagoon, being justified by the spatial configuration of the lagoons and the assumptions adopted in the model. The regions with the greatest tendency for metal accumulation were sediment, soil and road dust. The greater degree of complexity of the model compared to a more simplified model and the greater quantity/reliability of the environmental data increased the accuracy of the model s predictions. These results certify the use of this type of modeling as an auxiliary tool in the control, classification and risk assessment, and decision-making by the competent agencies.

Die Gletscherschmelze in den südlichen Anden trägt maßgeblich zum Anstieg des Meeresspiegels der letzten Jahrzehnte bei und beeinflusst regional die saisonale Wasserverfügbarkeit. In jüngster Zeit wurde eine rapide Zunahme der Massenverluste insbesondere einzelner großer Auslassgletscher des Südlichen Patagonisches Eisfeldes beobachtet. Im Rahmen der Dissertation wurden die rezente Variabilität des Klimas und der klimatischen Massenbilanz für ausgewählte vergletscherte Gebiete in Patagonien und Feuerland untersucht. Die Verbesserung unseres Verständnisses über räumliche und zeitliche Muster der klimatischen Massenbilanz, ihrer atmosphärischen Antriebsfaktoren und ihres Einflusses auf das in jüngster Vergangenheit beobachtete individuelle Gletscherverhalten, sind weitere wichtige Ziele. Da die Klimavariabilität die Hauptursache für lokale Veränderungen in der Kryosphäre der südlichen Anden ist, wurden langjährige meteorologische Beobachtungen im Gebiet der Gran Campo Nevado-Eiskappe im südlichsten Patagonien im Hinblick auf räumliche und zeitliche Variabilität untersucht und der Einfluss mesoskaliger Wettermuster und Modi atmosphärischer Oszillationen auf die Ausprägung des Klimas analysiert. Darüber hinaus wurde die rezente Variabilität der klimatischen Massenbilanz für ausgewählte Gletscher in Südpatagonien und Feuerland durch die Implementierung des Energie- und Massenbilanzmodells COSIMA simuliert. Eine unterschiedliche Ausprägung der Oberflächenmassenbilanz und geodätischer Massenbilanz unterstreicht wie wichtig ein besseres Verständnis über die Prozesse der klimatischen Massenbilanz und Eisdynamik ist. Des Weiteren wurden Simulationen der klimatischen Massenbilanz eingesetzt, um eine ausgeglichene Massenbilanz für rezente und vergangene Ausdehnungen des Gletschers Schiaparelli abzuleiten. Ziel war es, eine modellgestützte Annäherung an die klimatischen Bedingungen während der Kleinen Eiszeit zu simulieren. / Glacier mass loss of the Southern Andes contributes largely to sea-level rise during recent decades and also affects the regional water availability. Despite the overall glacier retreat of most glaciers in Patagonia and Tierra del Fuego, a recent increase in mass loss of individual glaciers has been observed. The recent variability of climate and climatic mass balance for selected glaciated study sites in Patagonia and Tierra del Fuego are investigated in this thesis. Improving our understanding on the spatial and temporal variations of climatic mass balance processes, its atmospheric drivers, and their impact on the recently observed individual glacier behavior are further important aims. Since climate variability is the key driver of local changes in the cryosphere in the Southern Andes, a unique record of meteorological observations across the Gran Campo Nevado Ice Cap in Southernmost Patagonia was analyzed with regard to main climate features and the relationship between the in-situ observations, large-scale climate modes and mesoscale weather patterns. Furthermore, recent climatic mass balance variability was simulated for selected glaciers in Southern Patagonia and Tierra del Fuego by implementing the ’COupled Snow and Ice energy and MAss balance model’ COSIMA. Contrasting patterns of positive simulated annual climatic mass balance and clearly negative geodetic mass balance were found for two neighboring glaciers of the Southern Patagonia Icefield between 2000 and 2014. This highlights the importance of understanding of both, the climatic mass balance, and the ice-dynamical processes. Climatic mass balance simulations were further used to derive glacier steady-state conditions for recent and past glacier extents of Schiaparelli Glacier, aiming for a model-based approximation of climate conditions during the Little Ice Age.

Excess nitrogen (N) from terrestrial landscapes poses environmental challenges as it moves via surface runoff and groundwater flows into aquatic ecosystems. Managing and anticipating the environmental challenges associated with these altered N inputs from terrestrial to aquatic ecosystems requires a deep understanding of how N is biogeochemically transformed, retained, and/or transported in streams and rivers. Here, I used long-term data on surface stream and groundwater chemistry as well as discharge to determine the main sources of N and estimate the N mass balance of a 1.4 km boreal stream reach. The goal was to evaluate daily net uptake or production rates of different N forms (ammonium - NH4-N, nitrate NO3-N and dissolved organic N - DON) throughout the seasons, and assess physical and chemical factors that may drive changes in net processing. The mass balances analysis revealed distinct patterns in net uptake among N forms. Notably, there was clear evidence of NH4-N and DON uptake (removal) in the stream, while NO3-N processing patterns showed neither clear uptake nor production. Further, variation in net uptake for NH4-N and DON was positively related to stream DOC, DOC:DIN, and C:N ratios, indicating that carbon rich conditions promoted greater N demand in this ecosystem. By comparison, variations in net NO3-N uptake or production at the reach scale were only weakly correlated with these carbon rich conditions. Finally, I assessed these patterns within the nutrient processing domains (NPDs) framework to characterize the behavior/character of the study reach. Accordingly, during the open water season, the stream reach acted mostly as a consumer for both NH4-N and DON, while on many dates it acted as a weak enhancer for NO3-N. These findings contribute to the broader understanding of N dynamics in boreal stream ecosystems and emphasize the complex interplay among organic and inorganic N forms, carbon dynamics, and nutrient processing in these environments. This knowledge is crucial for effective environmental management and conservation efforts in the region.

No description available.

No description available.